Glucocorticoid-induced leucine zipper enhanced expression in dendritic cells is sufficient to drive regulatory T cells expansion in vivo.

This information is current as of December 7, 2014.

Glucocorticoid-Induced Leucine Zipper Enhanced Expression in Dendritic Cells Is Sufficient To Drive Regulatory T Cells Expansion In Vivo

J Immunol 2014; 193:5863-5872; Prepublished online 31 October 2014; doi: 10.4049/jimmunol.1400758 http://www.jimmunol.org/content/193/12/5863

Supplementary Material References Subscriptions Permissions Email Alerts

http://www.jimmunol.org/content/suppl/2014/10/31/jimmunol.140075 8.DCSupplemental.html This article cites 45 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/193/12/5863.full#ref-list-1 Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscriptions Submit copyright permission requests at: http://www.aai.org/ji/copyright.html Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/cgi/alerts/etoc

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 9650 Rockville Pike, Bethesda, MD 20814-3994. Copyright © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606.

Downloaded from http://www.jimmunol.org/ at Monash University on December 7, 2014

Joseph Calmette, Mehdi Ellouze, Thi Tran, Soumaya Karaki, Emilie Ronin, Francis Capel, Marc Pallardy, Françoise Bachelerie, Roman Krzysiek, Dominique Emilie, Géraldine Schlecht-Louf and Véronique Godot

The Journal of Immunology

Glucocorticoid-Induced Leucine Zipper Enhanced Expression in Dendritic Cells Is Sufficient To Drive Regulatory T Cells Expansion In Vivo Joseph Calmette,*,†,1 Mehdi Ellouze,*,†,1 Thi Tran,*,†,1 Soumaya Karaki,*,† Emilie Ronin,*,† Francis Capel,*,† Marc Pallardy,*,‡ Françoise Bachelerie,* Roman Krzysiek,*,† Dominique Emilie,*,† Ge´raldine Schlecht-Louf,*,† and Ve´ronique Godotx,{,‖

I

mmune homeostasis requires the equilibrium between responses against pathogens and tolerance toward self-Ags and innocuous environmental Ags. Regulatory T cells (Tregs) are recognized as essential contributors to this balance. Two main Treg subsets have been described: the thymically derived CD4+Foxp3+CD25+ Tregs (tTregs) that differentiate in the thymus, and the peripheral *INSERM U996, Clamart, 92 140, France; †Faculte´ de Me´decine, Universite´ ParisSud, Le Kremlin Biceˆtre, 94 275, France; ‡Faculte´ de Pharmacie, Universite´ ParisSud, Chatenay Malabry, 92 296, France; xINSERM U955, Eq16, Cre´teil, 94 000, France; {Vaccine Research Institute, Cre´teil, 94 000, France; and ‖Universite´ Paris Est, Faculte´ de Me´decine, Cre´teil, 94 000, France 1

J.C., M.E., and T.T. contributed equally to this work.

Received for publication March 24, 2014. Accepted for publication October 6, 2014. This work was supported by the Agence Nationale pour la Recherche (Grant 09GENO-030-01), the Fondation pour la Recherche Me´dicale, the Universite´ Paris-Sud (Poˆle de Recherche et d’Enseignement Supe´rieur UniverSud and Appel a` Projet Attractivite´ 2012), a doctoral fellowship from the Ecole Doctorale Innovation The´rapeutique (Grant ED 425 to J.C.), a doctoral fellowship from INSERM (to M.E.), a doctoral fellowship from the Re´gion Ile de France (to S.K.), and a one-year postdoctoral fellowship from the Agence Nationale pour la Recherche (Grant 09-GENO-030-01 to S.K.). Address correspondence and reprint requests to Prof. Ve´ronique Godot or Dr. Ge´raldine Schlecht-Louf, INSERM U955, Eq16, 51 Avenue du Marećhal de Lattre de Tassigny, Cre´teil, 94 010 Cedex, France (V.G.) or INSERM U996, 32 Rue des Carnets, Clamart, 92 140, France (G.S.-L.). E-mail addresses: [email protected] (V.G.) or [email protected] (G.S.-L.) The online version of this article contains supplemental material. Abbreviations used in this article: BM-DC, bone marrow–derived DC; cDC, conventional DC; ctrl, control littermate; DC, dendritic cell; DCreg, regulatory DC; Dex, dexamethasone; eGFP, enhanced GFP; GC, glucocorticoid; GILZ, GC-induced leucine zipper protein; MFI, mean of fluorescence intensity; MLN, mesenteric lymph node; pDC, plasmacytoid DC; polyI:C, polyinosinic:polycytidylic acid; pTreg, peripheral Treg; SDLN, skin draining lymph node; Tconv, conventional CD4 T cell; Treg, regulatory T cell; tTreg, thymically derived CD4+Foxp3+CD25+ Treg. Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1400758

Tregs (pTregs) that differentiate from naive conventional CD4 T cells (Tconvs) and acquire Foxp3 expression in the periphery (1). Tregs depend on dendritic cells (DCs) for both their differentiation and activation. Indeed, DCs are required to activate tTregs in periphery and to convert Tconvs into pTregs (2). The exact nature of the regulatory DCs (DCregs) that are endowed with Treg induction and activation capacity is not entirely understood. Convergent studies show that this activation is not a default pathway but rather an active process that requires the expression of a regulatory genetic program by specialized DC subsets (3). Both developmental characteristics and environmental signals are reported to impact tolerogenic potential of DCs. IL-10, TGF-b, glucocorticoids (GCs), vasoactive intestinal peptide, vitamin D3, antioxidative vitamins, and retinoic acid, used alone or in combination, skew DC maturation toward DCregs (4, 5). However, the molecular mechanisms controlling the maturation of DCs into DCregs remain elusive. The GC-Induced Leucine Zipper protein (GILZ) is a GC-induced protein initially reported in dexamethasone (Dex)-treated thymocytes (6). We have previously shown ex vivo that the level of GILZ expression in human DCs controls their tolerance-inducing function (7–9) and extended to other hematopoietic cells these observations (10, 11), which were confirmed by others (12–15). GILZ expression is partly dependent on endogenous steroids in vivo (12), and its expression in DCs is enhanced in vitro by GC, IL-10, TGF-b (7, 8), mitomycin C (16), rapamycin (15), vitamin D3 (15), and tumor environment (17). Human GILZhi-DCs secrete large amounts of IL-10 compared with GILZlow-DCs and by this way induce Ag-specific CTLA-4+IL-10+ CD4+ Tregs, some of which coexpress Foxp3 (7, 8). Furthermore, we have shown in humans that GILZ mediates the immunoregulatory effects of GC


Tolerance induction by dendritic cells (DCs) is, in part, mediated by the activation of regulatory T cells (Tregs). We have previously shown in vitro that human DCs treated with glucocorticoids (GCs), IL-10, or TGF-b upregulate the GC-Induced Leucine Zipper protein (GILZ). GILZ overexpression promotes DC differentiation into regulatory cells that generate IL-10– producing Ag-specific Tregs. To investigate whether these observations extend in vivo, we have generated CD11c-GILZhi transgenic mice. DCs from these mice constitutively overexpress GILZ to levels observed in GC-treated wild-type DCs. In this article, we establish that GILZhi DCs display an accumulation of Foxp3+ Tregs in the spleens of young CD11c-GILZhi mice. In addition, we show that GILZhi DCs strongly increase the Treg pool in central and peripheral lymphoid organs of aged animals. Upon adoptive transfer to wild-type recipient mice, OVA-loaded GILZhi bone marrow–derived DCs induce a reduced activation and proliferation of OVA-specific T cells as compared with control bone marrow–derived DCs, associated with an expansion of thymus-derived CD25+Foxp3+ CD4 T cells. Transferred OVA-loaded GILZhi DCs produce significantly higher levels of IL-10 and express reduced levels of MHC class II molecules as compared with OVA-loaded control DCs, emphasizing the regulatory phenotype of GILZhi DCs in vivo. Thus, our work demonstrates in vivo that the GILZ overexpression alone is sufficient to promote a tolerogenic mode of function in DCs. The Journal of Immunology, 2014, 193: 5863–5872.

5864

GILZ-EXPRESSING DENDRITIC CELLS EXPAND Tregs IN VIVO Cell Tracer Kit; Invitrogen) during 15 min at room temperature. CD4 T cells were either injected i.v. into CD45.1 C57BL/6 mice (5 3 106 cells/ mouse) or used in vitro.

Western blots Protein extracts from 2 3 106 BM-DCs or splenic DCs were lysed with 20 ml lysis buffer (Roche Applied Science) as previously described (8). Fifty micrograms of proteins was loaded on a 4–12% gradient polyacrylamide gel (NuPAGE; Invitrogen). GILZ staining was performed with a rabbit anti-GILZ Ab (FL-134 clone; Santa Cruz) and a donkey anti-rabbit secondary Ab conjugated to peroxidase (Amersham, GE-Healthcare) as described in Hamdi et al. (8). b-Actin (eBioscience SAS) expression was used as loading control.

Ag capture and processing

Materials and Methods

A total of 2 3 105 DCs was distributed in wells of a 96-well plate. Ag capture was measured using Alexa 488–labeled OVA (Molecular Probes), dextran-FITC, and zymosan-FITC (both from Sigma-Aldrich). Dextran polymers are preferentially taken up by mannose receptor (CD206)–dependent endocytosis (22), OVA by both CD206-dependent endocytosis and macropinocytosis (23, 24), and zymosan by dectin-1–dependent endocytosis (25). OVA DQ (Molecular Probes) was used for Ag processing assay (26). After 90 min or 5 h of incubation at 4˚C or 37˚C, DCs were labeled with anti-CD11c Abs and analyzed on FACSCalibur or LSR Fortessa (BD Biosciences, San Diego, CA). D means of fluorescence intensity (MFIs) were calculated by subtracting the MFI at 4˚C (control) from the MFI at 37˚C.

Mice

DC activation for phenotype and cytokine/chemokine detection

CD11c-GILZhi mice were generated by a single targeted insertion (at the hprt docking site on the X chromosome) of GILZ coding sequences driven by the CD11c promoter (Nucleis, Lyon, France; Supplemental Fig. 1). Genetic uniformity was reached by a 10-generation backcross to the C57BL/6J background (.98%). The resulting CD11c-GILZhi transgenic mice are viable and fertile, with litters of normal size and sex ratio. Mice were maintained by heterozygous crossing with C57BL/6 males and CD11c-GILZhi mice were compared with controls from the same litter. Unless specified, hemizygous male mice were used between 8 and 14 wk of age. OT-II TCR transgenic mice were obtained from Charles River (L’Arbesle, France) and CD45.1 C57BL/6 mice from CDTA (Orleáns, France). Foxp3enhanced GFP (eGFP) and IL-10–GFP mice were kind gifts from Dr. B. Malissen (Centre d’Immunologie de Marseille-Luminy, Marseille, France) (18) and Dr. R. Flavell (Yale University School of Medicine, New Haven, CT) (19), respectively. Mice were bred in the Animex animal facility under specific pathogenfree facilities. Experimentation was approved by the local Ethics Committee for Animals (C2EA-26; Animal Care and Use Committee, Villejuif, France) and complied with French and European guidelines for the use of laboratory animals.

Proteins, peptides, and activators The synthetic OVA323–339 (ISQAVHAAHAEINEAGR) peptide corresponding to the immunodominant I-Ab–restricted T cell epitope from OVA was purchased from ANASPec (Fremont, CA). The OVA protein was obtained from Sigma-Aldrich. TLR agonists were purchased from Invivogen (Cayla SAS).

RNA extraction and quantitative RT-PCR RNAs were extracted from 2 to 5 3 106 bone marrow–derived DCs (BM-DCs) according to manufacturer’s instructions (RNeasy Plus Micro kit; Qiagen, France). cDNA were obtained by reverse transcription with the MuLV retrotranscriptase (Applied Biosystems). Quantitative RT-PCR was performed using SYBR Green I Master Mix kit (Roche Applied Science) and LightCycler 480 (Roche). Normalization was made by standard curve method based on b-actin expression. The primers sequences are available upon request.

Cell preparations BM-DCs were generated as previously described (20), using DC culture medium (CM; Life Technologies; RPMI 1640, 5% FCS, 50 mg/ml gentamicin, 5 3 1025 M 2-ME, and 20 ng/ml murine rGM-CSF; PreproTech). Cells (containing .90% of CD11c+ cells) were used at days 9 and 10. Splenic DCs were recovered from CD11c-GILZhi or littermate mice as described by Schlecht et al. (21). Magnetic cell sorting was done using autoMACS (Miltenyi Biotec). Cell purity ranged from 90 to 95%. CD4 T cells were purified from CD45.2 OT-II mice by negative magnetic selection using CD4 T Cell Isolation kit (autoMACS; Miltenyi Biotec). A biotinylated anti-CD25 Ab was added to the Ab mixture for Foxp3+CD25+ T cell depletion. T cells were labeled with 5 mM CFSE (Vybrant CFDA SE

The phenotype and cytokine/chemokine secretion profiles of DCs were assessed by incubating 2 3 105 BM-DCs in 96-well plates with 1.5 mM CpG (TLR9 agonist), 1 mg/ml LPS (TLR4 agonist), 10 mg/ml polyinosinic:polycytidylic acid (polyI:C; TLR3 agonist), or 10 mg/ml zymosan (TLR2 agonist). After 15 h, culture supernatants were recovered for cytokine measure, and DC phenotype was analyzed by flow cytometry. Cytokines and chemokines were assessed by ELISA using IL-10 and IL12(p70) BD OptEIA Set Mouse (BD Biosciences) or by Luminex assay using the Procarta mouse 26-plex from eBioscience.

T cell stimulation in vitro and in vivo BM-DCs were loaded with OVA or OVA323–339 peptide for 5 h and 90 min, respectively, at 37˚C. In in vitro assays, 2.5 3 103, 5 3 103, or 104 DCs per well were used in triplicate to stimulate 5 3 104 CFSE-labeled OT-II CD4 T cells. After 72 h of incubation, T cell proliferation was analyzed by flow cytometry. Splenic DCs were loaded with OVA for 4 h and cocultured with OT-II CD4 T cells at a 1:10 ratio. After 72 h, [3H]thymidine (50 mCi/well) was added to cultures. Cells were harvested 6 h later to evaluate incorporated [3H]thymidine by scintillation counting. In in vivo assays, 5 3 105 BM-DCs were transferred i.v. into CD45.1 mice 1 d after OT-II T cell injection. Sixty-six hours later, CD45.1 mice were sacrificed and spleens recovered as described in Schlecht et al. (27). T cell phenotype and proliferation were analyzed by flow cytometry.

Flow cytometry Cells suspended at 107 cells/ml in PBS-FCS 2% were first incubated with anti-CD16/32 (2.4G2 clone) Fc-Block, then labeled for 15–30 min at 4˚C in the dark with combinations of mAbs (clone): anti-CD11c (HL-3), -B220 (RA3-6B2), -CD45RA (14.8), -CD11b (M1/70), -CD8a (53-6.7), -CD40 (HM40-3), -CD86 (GL1), -CD3 (17A2), -CD4 (RM4-5), -CD8 (53-6.7), -CD62L (MEL-14), -CD44 (IM7), -CD11b (M1/70), -CD25 (PC61), -MHCII (M5/114.15.2), –PD-L1 (MIH5), –PD-L2 (TY-25), -ICOSL (HK5.3), -OX40 (OX86), –H-2Kb (CTKb), -CD45.2 (104), -CD27 (LG.7F9), -ICOS (7E.17G9), and -NK1.1 (PK136). mAbs and isotype controls were purchased from BD Biosciences (San Diego, CA), eBioscience (San Diego, CA), Biolegend (San Diego, CA), or Caltag (Buckingham, U.K.). For intracellular Foxp3 staining, Foxp3/Transcription Factor Staining Buffer Set (eBioscience, San Diego, CA) was used. For anti–IL-10 and anti– IFN-g staining, 106 splenocytes stimulated for 5 h with 1 mg/ml OVA323–339 peptide at 37˚C in the presence of GolgiPlug (BD Biosciences) were permeabilized using the Fix/Perm and Perm/Wash kit (BD Biosciences). Fluorescence was acquired on a FACSAria or a LSR Fortessa (BD Biosciences) flow cytometer and analyzed using the FlowJo Software (Tree Star).

Statistical analysis Statistical analysis of results was performed with GraphPad Prism software. Depending on experimental settings, (un)paired t test, one- or two-way ANOVA, or Wilcoxon signed rank test was used. The p values obtained


on DCs (8, 9). Downregulation of GILZ in DCs promotes the activation of effector CD4+ T cells (8, 13, 14, 17). To gain insight into the impact of GILZ on the regulatory functions of DCs in vivo, we generated CD11c-GILZhi transgenic mice. DCs from these mice constitutively overexpress GILZ to the same extent as GC-treated wild type DCs. In in vitro studies, GILZhi DC phenotype and function is not different from that observed in DCs from control littermates (ctrl). However, CD11cGILZhi mice display an age-dependent accumulation of Foxp3+ Tregs in vivo. In adoptive transfer models, we further demonstrate that the GILZhi DCs acquire a regulatory phenotype after cognate interaction with T cells. They express reduced levels of MHC class II molecules, secrete higher amounts of IL-10, poorly induce Tconv proliferation and activation, but promote tTreg expansion. In conclusion, our work formally identifies GILZ as a factor that contributes to DC tolerogenic function in vivo and validates the CD11c-GILZhi mice as a relevant model for studying in vivo the therapeutic potential of GILZ-expressing DCregs in immunopathologies relying on Treg defects.


5865

from multiple comparisons were adjusted by Bonferroni correction. The 5% threshold was considered significant. *p , 0.05, **p , 0.01.

Results Mice with constitutive overexpression of GILZ in DCs display normal lymphocyte and DC distribution in lymphoid tissues

GILZhi DC phenotype in vitro

hi

Increased Treg pool in CD11c-GILZ mice Because human monocyte-derived DCs overexpressing GILZ induce Treg differentiation in vitro (7–9), we investigated whether GILZ overexpression in DCs modified the Treg compartment in vivo. To address this issue, we first traced Foxp3-eGFP CD4+

The basal phenotype of DCs was investigated on either freshly isolated splenic DCs or BM-DCs. In both cases, the expression levels of CD80, CD86, CD40, MHC class I, or MHC class II molecules were similar between CD11c-GILZhi and ctrl mice (Fig. 3A, 3B). We further looked at the expression of several

FIGURE 1. GILZ is constitutively and selectively overexpressed in DCs from CD11c-GILZhi mice. (A and B) Quantification of GILZ expression in freshly isolated splenic DCs or BM-DCs derived from CD11c-GILZhi or ctrl mice (ctrl). BM-DCs were treated or not with 1027 M Dex for 18 h. (A) Proteins were extracted to quantify GILZ expression by Western blot. A protein extract from a Dex-stimulated mast cell line HMC-1 is used as positive control (pos). After stripping, the membrane was reprobed for b-actin (42 kDa, lower row). Results are representative of three experiments. (B) GILZ mRNA expression in splenic DCs or BM-DCs was determined by real-time quantitative RT-PCR (qRT-PCR) and normalized over b-actin mRNA expression. Results correspond to the mean 6 SEM of three to four independent experiments. (C–E) GILZ expression in lymphocytes and NK cells from CD11c-GILZhi mice. CD4 T cells, CD8 T cells were FACS-sorted from spleens, MLNs, and SDLNs from GILZhi or ctrl mice and NK cells from liver and spleen. GILZ mRNA levels were determined by real-time qRT-PCR as in B. Mean 6 SEM of three to four mice per group. (F) DC distribution assessed by flow cytometry in CD11c-GILZhi (GILZhi) and littermate control (ctrl) mouse spleens. The percentages of conventional DCs (CD11chipDCA-12) and plasmacytoid DCs (CD11clopDCA-1+) are indicated in the dot plots. **p , 0.01.


In the CD11c-GILZhi transgenic mice, we used the CD11c promoter to drive enforced expression of GILZ in DCs (28). We confirmed an overexpression of GILZ mRNA and protein in splenic DCs and BM-DCs from CD11c-GILZhi mice as compared with their ctrl (Fig. 1A, 1B). Therefore, DCs from CD11c-GILZhi mice will be referred to as GILZhi DCs throughout the text. The enforced expression of GILZ driven by the CD11c promoter in transgenic mice and the expression of GILZ induced by Dex in ctrl mice were comparable, thereby confirming a regular level of GILZ in CD11c-GILZhi mice (Fig. 1A, 1B). Likewise, the level of GILZ expression was slightly increased in GILZhi BM-DCs after Dex treatment (Fig. 1A, 1B), pointing out a suboptimal level of GILZ in these transgenic cells. GILZ overexpression was restricted to DCs because no alteration in GILZ mRNA levels was observed in other hematopoietic cells such as CD4 T cells, CD8 T cells, splenic NK cells, or liver NK cells, a population known to be enriched in CD11c-expressing cells (29) (Fig. 1C–E). No change in the architecture (data not shown) and total cell numbers in spleen, skin draining lymph nodes (SDLNs), and mesenteric lymph nodes (MLNs) was observed in CD11c-GILZhi mice as compared with ctrl (Table I). Furthermore, the frequencies and absolute numbers of B cells, T cells, conventional DCs (cDCs), and plasmacytoid DCs (pDCs) were similar in secondary lymphoid organs of CD11c-GILZhi and ctrl littermates (Fig. 1F and Table I).

Tregs in several lymphoid organs of young (8–12 wk old) and aged (18 mo) Foxp3-eGFP 3 CD11c-GILZhi female mice obtained by crossing CD11c-GILZhi mice with Foxp3-eGFP reporter mice. We observed an increase of Foxp3+CD25+ Treg frequency restricted to the spleen in young Foxp3-eGFP 3 CD11c-GILZhi mice and generalized to the thymus and the secondary lymphoid organs for both CD25+ and CD252 Tregs in aged Foxp3-eGFP 3 CD11c-GILZhi mice as compared with Foxp3-eGFP ctrl littermates (Fig. 2A, 2B). Because both the CD11c-GILZ transgene and the Foxp3 gene are carried on X chromosome, subjected to X-inactivation, we checked GILZ overexpression in DCs from Foxp3-eGFP 3 CD11c-GILZhi females (Fig. 2C) and verified that the proportion of Foxp3-eGFP Tregs corresponded to half the Tregs detected by intracellular labeling as expected (Fig. 2D). Likewise, we confirmed the expansion of Tregs in lymphoid organs of young and 24-mo-old CD11c-GILZhi male by Foxp3 intracellular staining (Fig. 2E, 2F). However, no difference in the percentages and numbers of CD4+Foxp32 Tconvs was observed between ctrl and CD11c-GILZhi mice, whatever their age (Fig. 2E, 2F). The Treg phenotype analyzed in aged CD11c-GILZhi mice was similar to that of ctrl littermates for CD69, CTLA4, GITR, CD103, PD1, and CCR7 (data not shown). The thymus and SDLN of aged CD11c-GILZhi mice showed an increase of the ICOShi Treg subset (Fig. 2H), a subset reported to be highly suppressive and biased toward IL-10 production (30–32). This increase of ICOShi Tregs was also confirmed in young CD11c-GILZhi mice (Fig. 2G). Thus, the sole GILZ overexpression in DCs clearly promotes Treg accumulation in vivo, and this accumulation is associated with an increase in the ICOShi Treg subset.

5866

GILZ-EXPRESSING DENDRITIC CELLS EXPAND Tregs IN VIVO Table I.

DC, B cell, and T cell frequency in secondary lymphoid organs Organ Spleen Ctrl

Cell number (3106) cDCs, % pDCs, % B cells, % CD4 T cells, % CD8 T cells, %

100.4 1.43 0.37 50.7 36.3 14.3

6 6 6 6 6 6

SDLN GILZ

5.3 0.1 0.16 2.9 6.7 3.9

118 1.48 0.37 52.9 34 14

6 6 6 6 6 6

hi

7.5 0.3 0.25 6.1 2.7 3

Ctrl

4 0.8 0.26 31.1 48.5 23.6

6 6 6 6 6 6

MLN GILZ

0.37 0.1 0.01 3.5 8.4 13

5.2 1 0.22 29.5 44.6 23.4

6 6 6 6 6 6

hi

0.75 0.2 0.1 2.5 4.2 11

GILZhi

Ctrl

13 2.1 0.25 27.7 46.5 24.4

6 6 6 6 6 6

2.2 0.6 0.1 3.2 3.4 10

18.3 2.3 0.18 24.7 48.6 20.5

6 6 6 6 6 6

3.6 0.9 0.1 8.9 7.9 9.7

Spleen, SDLNs, and MLNs were harvested and single-cell suspensions prepared. The cells were counted, and the proportions of cDCs (CD11chi PDCA-12), pDCs (CD11clo PDCA-1+), B cells (CD19+B220+), CD4 T cells (CD3+CD4+), and CD8 T cells (CD3+CD8+) were determined by flow cytometry. Results are from six to eight mice per group.

FIGURE 2. Treg accumulation in CD11c-GILZhi mice. (A and B) Foxp3-eGFP 3 CD11c-GILZhi female mice were obtained by crossing transgenic CD11c-GILZhi female mice with Foxp3-eGFP male mice. Thymuses, spleens, and SDLNs from young (8–12 wk old) (A; n = 6–8 mice/group) and aged (18-mo-old) (B; n = 5–6 mice/group) Foxp3-eGFP 3 CD11c-GILZhi or Foxp3-eGFP littermate control mice were analyzed by flow cytometry. T cells were gated as single CD3 CD4 T cells and assessed for Foxp3 and CD25 expression. (C) GILZ overexpression in splenic DCs from Foxp3-eGFP 3 CD11c-GILZhi females was confirmed by quantitative RT-PCR (n = 5–6 mice/group). (D) Treg detection in the spleen of Foxp3-eGFP 3 CD11c-GILZhi and littermate control Foxp3-eGFP mice using eGFP signal (left) or intracellular Foxp3 staining (right) was compared to confirm the proportionality between the two detection systems (n = 4–5 mice/group). (E and F) Thymuses and spleens from young (E) or 24-moold (F) CD11c-GILZhi and control male mice were recovered. Frequencies and numbers of CD3+CD4+ Foxp3+ CD25+ or CD252 Tregs (left) and CD3+CD4+Foxp32 Tconv (right) are shown (n = 3–6 mice/group). (G and H) Frequency of ICOShiFoxp3+ T cells in lymphoid organs of young (G; n = 6–8 mice) and aged (H; n = 5–6 mice) Foxp3-eGFP 3 CD11cGILZhi or Foxp3-eGFP littermate female mice are shown. Mean 6 SEM, two-tailed unpaired t test. *p , 0.05, *p , 0.01.

freshly isolated DCs or BM-DCs does not significantly modify their basal phenotype ex vivo and in vitro. We next addressed the impact of GILZ overexpression on DC maturation in vitro. Analysis of the expression of CD86, CD40, and


membrane markers for tolerogenic DCs. The results showed no increased expression of PD-L1, PD-L2, ICOSL, and OX40L on GILZhi splenic DCs or GILZhi BM-DCs as compared with their controls (Fig. 3A, 3B). Thus, the sole GILZ overexpression in


5867 viral transduction display deficient OVA uptake (17), we first tested the capacity of GILZhi splenic DCs and BM-DCs to internalize and degrade soluble (OVA, dextran) and particulate (zymosan) Ags. No differences in dextran or zymosan capture were observed (Fig. 4A). Consistently with the results published by Lebson et al. (17), GILZhi BM-DCs displayed a reduced OVA uptake as compared with control BM-DCs after a short incubation period with this Ag, and OVA degradation was similarly reduced (Fig. 4A). However, these capture and degradation defects were transient and disappeared completely at later times (i.e., 5 h of incubation; Fig. 4B), and were not observed for splenic DCs (Fig. 4C). To test Ag presentation capacity, we loaded GILZhi and ctrl BM-DCs with OVA for 5 h to avoid differences in Ag access, and incubated them with CFSE-labeled OT-II CD4 T cells. As shown in Fig. 4D, GILZhi DCs induced OT-II cell proliferation as efficiently as ctrl BM-DCs. Similar results were obtained with splenic DCs (Fig. 4E). Overall, these data demonstrated that GILZhi DCs display no major alteration in their capacity to take up and present Ags in vitro.

Ag capture and presentation by GILZhi DCs in vitro

GILZhi DCs poorly activate effector T cells but promote tTreg expansion in vivo

The absence of a tolerant phenotype for GILZhi DCs in vitro prompted us to verify whether these DCs loaded with OVA were able to activate OVA-specific OT-II Tregs after in vivo adoptive transfer. Because murine DCs overexpressing GILZ upon lenti-

We next investigated the ability of OVA-loaded GILZhi DCs to prime naive Ag-specific T cells in vivo, and determined whether they favor the expansion of Ag-specific Tregs. To this end, CFSE-labeled

FIGURE 3. GILZhi DC phenotype and cytokine secretion at resting state or after TLR engagement. (A and B) DCs were harvested from the spleens (A) or derived from the bone marrow (B) of CD11c-GILZhi (GILZhi) and littermate control (ctrl) mice and analyzed by flow cytometry. DCs were gated as CD11c+ cells, and their expression of MHC class II molecules, costimulatory molecules, or inhibitory molecules was assessed. MFI and percents of positive cells as compared with isotype controls are indicated in regular and bold font, respectively. Results shown are representative of three experiments. (C) GILZhi and ctrl BM-DCs phenotype after stimulation for 15 h with TLR2 (zymosan), TLR3 (polyI:C), TLR4 (LPS) or TLR9 (CpG ODN) agonists, or Dex. MHC II molecules, CD40, and CD86 expression were measured by FACS on CD11c+ cells. Means 6 SEM of MFI fold increase for five to eight mice in each group are shown. (D) IL-10 and IL-12p70 concentrations were measured by ELISA in the culture supernatants from GILZhi and ctrl BM-DCs prepared in (C). Wilcoxon signed rank test corrected by Bonferroni for multiple comparison was used to compare treated ctrl or GILZhi DCs with unstimulated ctrl or GILZhi DCs, respectively. ***p , 0.005, **p , 0.01, *p , 0.05.


MHC class II molecules on GILZhi and ctrl BM-DCs after stimulation with TLR2 (zymosan), TLR3 (polyI:C), TLR4 (LPS), and TLR9 (CpG) ligands revealed no statistically significant differences. Because Dex downregulated costimulatory molecules on human DCs through GILZ induction (7, 8), we evaluated the expression of MHC class II molecules, CD40, and CD86 on Dextreated-GILZhi and ctrl DCs. An equivalent downregulation of MHC and costimulatory molecules was observed in both DC subsets (Fig. 3C). This might be explained by the higher expression of GILZ reached in vitro after DC treatment with Dex (see Fig. 1A, 1B) or by GILZ-independent effects of Dex on DC phenotype in mice. No differences in IL-10 or IL-12p70 production could be observed between GILZhi and control BM-DCs either in the steadystate or after activation (Fig. 3D). Other cytokines and chemokines were not differently expressed either (Supplemental Fig. 2 and data not shown). Taken together, these data indicate that GILZhi BM-DCs display no alteration of surface phenotype and cytokine production in vitro.

5868

GILZ-EXPRESSING DENDRITIC CELLS EXPAND Tregs IN VIVO

OT-II T cells (CD45.2) and GILZhi or control DCs (CD45.2) loaded with OVA were i.v. injected to CD45.1 congenic recipient mice (Fig. 5A). Three days after DC injection, spleens were recovered and CD45.2 OT-II T cells were analyzed by flow cytometry. CD4 T cells triggered by OVA-loaded GILZhi DCs proliferated less than those stimulated with control DCs (Fig. 5B, 5D). This decreased proliferation capacity was associated with a lower OT-II cell activation state as revealed by CD44 and CD62L staining (Fig. 5E, 5F). Similar results were obtained when DCs were loaded with the OVA323–339 peptide, further demonstrating that the difference in T cell activation between GILZhi and ctrl DCs is not due to a defect in Ag processing (Fig. 5B, 5D–F). The analysis of cytokine production by splenic OT-II T cells after a 5-h in vitro restimulation with OVA323–339 peptide revealed that OVA-loaded GILZhi DCs induced less IFN-g–producing T cells and more IL-10–producing T cells than ctrl DCs (Fig. 5G). Thus, although GILZhi DCs were as efficient as control DCs in inducing T cell proliferation in vitro, they were significantly less efficient in inducing T cell stimulation and proliferation in vivo. To test whether this reduced stimulation capacity was associated with the proliferation of Tregs, we performed a Foxp3 intracellular staining on the OT-II T cells recovered after the adoptive transfer. This experiment revealed a significantly higher proportion of CD4+Foxp3+CD25+ Tregs after T cell priming with GILZhi DCs (Fig. 5H, 5J). These Tregs were mainly found among dividing cells (Fig. 5I), suggesting that the increase of Tregs observed after OVA-loaded GILZhi DC injection results from Treg proliferation. To determine whether these proliferating Tregs derived from tTregs or were differentiated from Foxp32 Tconv, we reproduced this experiment using OT-II T cells depleted from CD25+ T cells,

which correspond essentially to Tregs in naive OT-II mice. Although CD25-depleted OT-II T cells primed by OVA-loaded GILZhi DCs still proliferate far less than those activated by OVA-loaded ctrl DCs (Fig. 5C, 5D) and display a less activated phenotype (Fig. 5E, 5F), no Treg accumulation was observed in mice that had received OVA-loaded GILZhi DCs (Fig. 5J). These adoptive transfers formally demonstrated that GILZhi DCs are poor inducers of CD4 T cell proliferation and activation, but that they preferentially support tTreg expansion in vivo. GILZhi DCs display a regulatory phenotype in vivo To shed light on the apparent paradoxical phenotype displayed by GILZhi DCs in vitro and in vivo, where they show no specific phenotype or cytokine secretion, respectively, (Fig. 3) and tolerogenic properties, we analyzed GILZhi DC phenotype in vivo after an encounter with T cells. We were particularly interested in their production of IL-10 because we previously demonstrated that one characteristic of tolerant human GILZhi DCs was a higher IL10 production. Therefore, we crossed the CD11c-GILZhi and ctrl mice with IL-10–GFP reporter mice (19) and reproduced the earlier-described adoptive transfer in CD45.1 congenic recipient mice, using GILZhi or ctrl DCs coexpressing GFP with IL-10 (Fig. 6A). Equivalent numbers of DCs were recovered in each condition (data not shown), showing that the differences in T cell priming by GILZhi DCs cannot be assigned to deficient homing to the spleen. Although no difference was observed for CD86 expression on GILZhi and ctrl DCs (Fig. 6B), the analysis of MHC class II molecules expression revealed a significant decrease on OVA-loaded GILZhi DCs (Fig. 6C). OVA-loaded GILZhi DCs also produced significantly more IL-10 than ctrl OVA-loaded DCs


FIGURE 4. Ag uptake, processing, and presentation by GILZhi BM-DCs and splenic DCs in vitro. BM-DCs (A and B) or splenic DCs (C) from CD11cGILZhi (GILZhi) and littermate control (ctrl) mice were incubated with OVA–Alexa 488, OVA-DQ, dextran-FITC, or Zymosan-FITC at 4˚C or 37˚C, for 90 min (A and C) or 5 h (B). After washing the cells, Ag capture and processing were measured by flow cytometry. Percents of positive cells and D MFI (MFI at 37˚C 2 MFI at 4˚C) are indicated as means 6 SEM for 5–12 mice per group. Two-tailed unpaired t test was used to compare GILZhi DCs with ctrl DCs. **p , 0.01, *p , 0.05. (D) GILZhi and ctrl BM-DCs were loaded with OVA (1 mg/ml) for 5 h and used to stimulate CFSE-labeled OT-II CD4 T cells in vitro. DC/OT-II cell ratio was 1:5, one representative experiment out of four. (E) Splenic DCs were loaded for 4 h with OVA (1 mg/ml) and cocultured for 72 h with OT-II cells at a 1:10 ratio. Results correspond to means 6 SEM of [3H]thymidine incorporation (n = 4).


5869

(Fig. 6D). Using DCs from CD11c-GILZhi 3 IL-10–GFP and ctrl 3 IL-10–GFP mice, we confirmed that the DCs used for the adoptive transfer displayed the same capacity to produce IL-10 in vitro upon TLR2 engagement (Fig. 6E), results consistent with those obtained in Fig. 3D. The significant increase in IL-10 production in vivo was observed only for OVA-loaded GILZhi DCs, and not for nonloaded GILZhi DCs, suggesting a role for DC–T cell interactions in DC phenotype modification. However, the sole engagement of CD40 on GILZhi DCs by anti-CD40 Abs was not sufficient to increase their IL-10 production in vitro (data not shown). The analysis of the CD45.2+CD4+ OT-II T cells recovered from the spleens 15 h after DC transfer revealed a similar activation of T cells that had interacted with OVA-loaded GILZhi and ctrl DCs, as assessed by CD25 and CD69 staining (Fig. 6F). Overall, these data show that upon in vivo transfer, GILZhi DCs acquire a regulatory phenotype, characterized by a reduced MHC class II expression at the cell surface and an increased IL-10 production.

Discussion We set up a new mouse strain that constitutively expresses enhanced GILZ levels in DCs to investigate in vivo the contribution of GILZ to the tolerogenic function of DCs. In this study, we show that

the sole GILZ overexpression in DCs is sufficient to promote thymus-derived Foxp3+ Treg expansion in vivo. We previously reported that high levels of GILZ expression in human DCs endow these cells with a regulatory potential, stimulating the proliferation of IL-10–secreting Ag-specific Tregs in vitro (7–9). In this article, we show that young CD11c-GILZhi mice have significantly increased frequencies of Tregs in the spleen. Noteworthy, this increase in Treg frequency is exacerbated and generalized to all the lymphoid organs in 18- to 24-mo-old animals (33, 34), pointing to an age-dependent accumulation of Tregs driven by GILZ-expressing DCs. In addition, the pool of ICOS-expressing Tregs is significantly increased in CD11c-GILZhi mice. ICOS contributes to the homeostatic survival of Tregs and to the expansion of Ag-specific Tregs (31). The ICOShi Tregs were reported to be prone to produce IL-10 (30, 31, 35). The increase of the ICOShi Tregs in the CD11c-GILZhi mice may therefore result from a constitutively enhanced activation of Tregs that could participate in Treg accumulation over time. To exclude the possibility that the increase in Treg frequency observed in the CD11c-GILZhi mice results from a CD11c-driven expression of GILZ in a subset of T cells (36) that could promote their survival (37) or alter their differentiation (38), we used an


FIGURE 5. GILZhi DCs promote tTreg expansion in vivo. (A) CFSE-labeled OT-II CD4 T cells are injected to congenic CD45.1 mice the day before DC injection. Ctrl or GILZhi DCs are loaded with OVA or OVA323–339 and i.v. injected to recipient mice at day 0. Sixty-six hours later, spleens from recipient mice are recovered and analyzed by FACS for OT-II T cell proliferation (B–D, I), phenotype (E, F, H, and J), and cytokine secretion (G). OT-II cells are gated as CD45.2+CD3+CD4+ T cells. (B) Representative CFSE dilution profiles after in vivo activation of OT-II T cells by nonloaded DCs, OVA-loaded GILZhi and ctrl DCs, or OVA323–339–loaded GILZhi and ctrl DCs. (C) Representative CFSE dilution profiles after in vivo activation of CD25-depleted OT-II T cells (CD25-depl OT II) by nonloaded DCs and OVA-loaded GILZhi and ctrl DCs. (D) OT-II cell numbers recovered from mice receiving OVA- or OVA323–339–loaded GILZhi and ctrl DCs (OVA, n = 5; OVA323–339, n = 2–3; CD25-depl OT-II OVA, n = 2). (E and F) Percentages of CD44hi (E) and CD62Llo (F) cells among divided OT-II cells after in vivo stimulation by OVA- or OVA323–339–loaded DCs. (G) Ag-specific IFN-g and IL-10 secretion by OT-II T cells from mice injected with OVA-loaded DCs after a 5-h in vitro stimulation with OVA323–339 peptide in the presence of GolgiPlug. (H) Frequency of CD25+Foxp3+ Tregs among OT-II T cells assessed by Foxp3 intracellular staining. Representative dot plots for Foxp3 and CD25 staining on OT-II T cells are shown, with the percentages of CD25+Foxp3+ Tregs. (I) Representative CFSE profiles showing Foxp3+ CD4+CD45.2+CD3+ Treg proliferation. (J) Histograms present means 6 SEM of the percentages of CD25+Foxp3+ OT II Tregs after in vivo stimulation by OVA- or OVA323–339–loaded GILZhi and ctrl DCs (OVA, n = 5; OVA323–339, n = 2–3; CD25-depl OT-II OVA, n = 2). Two-way ANOVA followed by Bonferroni posttests were used to compare the results from mice injected with OVA-loaded GILZhi and ctrl DCs. **p , 0.01, * p , 0.05.

5870


adoptive transfer model in which only the transferred DCs overexpress GILZ. When OVA-loaded GILZhi DCs and OT-II T-cells were cotransferred to wild-type recipient mice, OVA-specific T cell proliferation and activation were reduced as compared with that induced by OVA-loaded control DCs. This was associated with an increase in the percentages of CD25+Foxp3+ Tregs and of IL-10–producing cells among proliferating cells, which was consistent with our findings in humans (8, 9). Because GILZhi DCs display a significant but transient reduction of OVA uptake and processing in vitro, we verified that the kinetics of loading we used in adoptive transfer results in similar OVA uptake by control and GILZhi DCs, suggesting that the reduced T cell activation by GILZhi DCs is not due to differences in Ag uptake capacity. This is further supported by the observation that control and GILZhi DCs induced a similar T cell proliferation in vitro, and by the experiments with OVA323–339 peptide–loaded DCs in vivo. One question raised by the Treg proliferation achieved in vivo in our adoptive transfer model, as well as in vitro with human GILZhi DCs (8, 9), is the nature of these Tregs. To determine whether GILZhiDC-expanded Tregs originate from pre-existing tTregs or from the conversion of Tconvs into pTregs, we used OT-II T cells depleted from Foxp3+CD25+ cells. This depletion completely abrogated Treg accumulation, formally demonstrating that GILZhi DCs trigger the proliferation of tTregs. In these experiments, OVA-specific CD4 Tconvs barely proliferate upon challenge with GILZhi DCs despite absence of tTreg amplification. This low proliferation may simply result from the reduced MHC II expression by OVA-loaded GILZhi DCs in vivo but could also reflect

a state of anergy as classically observed when a T cell meets its Ag in the absence of costimulation (39) or alternatively when it is activated in the presence of IL-10 (40). At this stage of their differentiation, it is not possible to distinguish anergic T cells induced by TCR ligation in the absence of costimulation from IL-10–anergized T cells, because both populations fail to produce cytokines (40, 41). IL-10 and TGF-b production by IL-10–anergized T cells require several rounds of restimulation (42). In this article, we report that OVA-loaded GILZhi DCs produce significantly higher levels of IL-10 and do not harbor reduced CD86 expression as compared with OVA-loaded control DCs. Thus, GILZhi DCs could contribute to the induction of IL-10–anergized T cells, a hypothesis supported by our previous observation in humans (8). Remarkably, the regulatory phenotype of GILZhi DCs was only observed in vivo and restricted to Ag-loaded DCs, suggesting that it may be acquired in vivo upon DC–T cell cognate interactions that are currently under investigation. This would explain why we did not detect a global phenotype alteration in DCs from GILZhi DCs in vitro under steady-state or after TLR engagement, results that corroborate the data obtained by other groups using GILZknockout mice (43, 44). GILZ is known to inhibit transcription factors like NF-kB, AP1, and FoxO3 (12). One possibility is that GILZ alters the signaling pathways activated in DCs upon their interaction with T cells in vivo. Our results do not support a role of GILZ in Dex-mediated effects on murine DC phenotype (reduction in MHC class II molecules and CD86 expression), as opposed to the results we observed with human monocyte-derived DCs (8). This discrepancy may result from the


FIGURE 6. GILZhi DCs acquire a regulatory phenotype upon in vivo transfer. (A) Adoptive transfer approach to analyze DC phenotype (B and C) and their IL-10 production (D) after T cell encounter in vivo. Transferred DCs were gated as CD45.2+CD11c+MHCII+ cells. (B) CD86 expression at the surface of the CD45.2 DCs (C), MHC class II expression, and (D) IL-10 production were assessed by FACS analysis. Left histograms show the results for one representative experiment out of three. MHC class II and IL-10 expression by CD45.1+ endogenous DCs (gray lines) and CD45.2+ transferred DCs (black lines) are presented. MFI and percents of positive cells as compared with endogenous DCs are indicated in regular and bold fonts, respectively. Right histograms represent the MFI 6 SEM for the three experiments. (E) IL-10 production by GILZhi 3 IL-10–GFP and ctrl IL-10–GFP DCs in the resting state (medium) and upon in vitro stimulation with zymosan (n = 3). Two-way ANOVA followed by Bonferroni posttests were used for statistical analysis. **p , 0.01, *p , 0.05. (F) CD45.2+CD3+CD4+ OT-II T cells activation was revealed by the percentages of cells expressing CD25 (left panel) and CD69 (right panel) (n = 3).

The Journal of Immunology experimental procedures used to upregulate GILZ in human DCs (Dex treatment and GILZ gene transfection) or intrinsic differences, which may exist between human and murine DCs. The latter hypothesis is supported by the fact that Dex treatment of murine DCs did not induce IL-10 production in contrast with human DCs, pointing to interspecies differences in responses of DCs to GC exposure. Collectively, our work demonstrates that DCs overexpressing GILZ acquire a tolerogenic function, associated with their ability to produce IL-10, and expand Foxp3+ Tregs in vivo. These results identify the CD11c-GILZhi mouse as a unique and robust model to investigate in vivo the therapeutic potential of GILZ modulation in DCs in immunopathologies such as autoimmune diseases (45) or allergies and in allograft rejection.

Acknowledgments

Disclosures The authors have no financial conflicts of interest.

References 1. Liston, A., and D. H. Gray. 2014. Homeostatic control of regulatory T cell diversity. Nat. Rev. Immunol. 14: 154–165. 2. Povoleri, G. A., C. Scotta`, E. A. Nova-Lamperti, S. John, G. Lombardi, and B. Afzali. 2013. Thymic versus induced regulatory T cells - who regulates the regulators? Front. Immunol. 4: 169. 3. Yamazaki, S., and A. Morita. 2013. Dendritic cells in the periphery control antigen-specific natural and induced regulatory T cells. Front. Immunol. 4: 151. 4. Maldonado, R. A., and U. H. von Andrian. 2010. How tolerogenic dendritic cells induce regulatory T cells. Adv. Immunol. 108: 111–165. 5. Reis e Sousa, C. 2006. Dendritic cells in a mature age. Nat. Rev. Immunol. 6: 476–483. 6. D’Adamio, F., O. Zollo, R. Moraca, E. Ayroldi, S. Bruscoli, A. Bartoli, L. Cannarile, G. Migliorati, and C. Riccardi. 1997. A new dexamethasoneinduced gene of the leucine zipper family protects T lymphocytes from TCR/ CD3-activated cell death. Immunity 7: 803–812. 7. Cohen, N., E. Mouly, H. Hamdi, M. C. Maillot, M. Pallardy, V. Godot, F. Capel, A. Balian, S. Naveau, P. Galanaud, et al. 2006. GILZ expression in human dendritic cells redirects their maturation and prevents antigen-specific T lymphocyte response. Blood 107: 2037–2044. 8. Hamdi, H., V. Godot, M. C. Maillot, M. V. Prejean, N. Cohen, R. Krzysiek, F. M. Lemoine, W. Zou, and D. Emilie. 2007. Induction of antigen-specific regulatory T lymphocytes by human dendritic cells expressing the glucocorticoidinduced leucine zipper. Blood 110: 211–219. 9. Karaki, S., G. Garcia, C. Tcherakian, F. Capel, T. Tran, M. Pallardy, M. Humbert, D. Emilie, and V. Godot. 2014. Enhanced glucocorticoid-induced leucine zipper in dendritic cells induces allergen-specific regulatory CD4(+) T-cells in respiratory allergies. Allergy 69: 624–631. 10. Berrebi, D., S. Bruscoli, N. Cohen, A. Foussat, G. Migliorati, L. BouchetDelbos, M. C. Maillot, A. Portier, J. Couderc, P. Galanaud, et al. 2003. Synthesis of glucocorticoid-induced leucine zipper (GILZ) by macrophages: an antiinflammatory and immunosuppressive mechanism shared by glucocorticoids and IL-10. Blood 101: 729–738. 11. Godot, V., G. Garcia, F. Capel, M. Arock, I. Durand-Gasselin, M. L. Asselin-Labat, D. Emilie, and M. Humbert. 2006. Dexamethasone and IL-10 stimulate glucocorticoidinduced leucine zipper synthesis by human mast cells. Allergy 61: 886–890. 12. Beaulieu, E., and E. F. Morand. 2011. Role of GILZ in immune regulation, glucocorticoid actions and rheumatoid arthritis. Nat. Rev. Rheumatol. 7: 340–348. 13. Benkhoucha, M., N. Molnarfi, I. Dunand-Sauthier, D. Merkler, G. Schneiter, S. Bruscoli, C. Riccardi, Y. Tabata, H. Funakoshi, T. Nakamura, et al. 2014. Hepatocyte Growth Factor Limits Autoimmune Neuroinflammation via Glucocorticoid-Induced Leucine Zipper Expression in Dendritic Cells. J. Immunol. 193: 2743–2752.

¨ . Met, and I. M. Svane. 2013. Silencing of the glucocorticoid14. Cathelin, D., O induced leucine zipper improves the immunogenicity of clinical-grade dendritic cells. Cytotherapy 15: 740–749. 15. Zimmer, A., J. Bouley, M. Le Mignon, E. Pliquet, S. Horiot, M. Turfkruyer, V. Baron-Bodo, F. Horak, E. Nony, A. Louise, et al. 2012. A regulatory dendritic cell signature correlates with the clinical efficacy of allergen-specific sublingual immunotherapy. J. Allergy Clin. Immunol. 129: 1020–1030. 16. Terness, P., T. Oelert, S. Ehser, J. J. Chuang, I. Lahdou, C. Kleist, F. Velten, G. J. Ha¨mmerling, B. Arnold, and G. Opelz. 2008. Mitomycin C-treated dendritic cells inactivate autoreactive T cells: toward the development of a tolerogenic vaccine in autoimmune diseases. Proc. Natl. Acad. Sci. USA 105: 18442– 18447. 17. Lebson, L., T. Wang, Q. Jiang, and K. A. Whartenby. 2011. Induction of the glucocorticoid-induced leucine zipper gene limits the efficacy of dendritic cell vaccines. Cancer Gene Ther. 18: 563–570. 18. Wang, Y., A. Kissenpfennig, M. Mingueneau, S. Richelme, P. Perrin, S. Chevrier, C. Genton, B. Lucas, J. P. DiSanto, H. Acha-Orbea, et al. 2008. Th2 lymphoproliferative disorder of LatY136F mutant mice unfolds independently of TCRMHC engagement and is insensitive to the action of Foxp3+ regulatory T cells. J. Immunol. 180: 1565–1575. 19. Kamanaka, M., S. T. Kim, Y. Y. Wan, F. S. Sutterwala, M. Lara-Tejero, J. E. Galań, E. Harhaj, and R. A. Flavell. 2006. Expression of interleukin-10 in intestinal lymphocytes detected by an interleukin-10 reporter knockin tiger mouse. Immunity 25: 941–952. 20. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu, and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor. J. Exp. Med. 176: 1693–1702. 21. Schlecht, G., C. Leclerc, and G. Dadaglio. 2001. Induction of CTL and nonpolarized Th cell responses by CD8alpha(+) and CD8alpha(-) dendritic cells. J. Immunol. 167: 4215–4221. 22. Sallusto, F., M. Cella, C. Danieli, and A. Lanzavecchia. 1995. Dendritic cells use macropinocytosis and the mannose receptor to concentrate macromolecules in the major histocompatibility complex class II compartment: downregulation by cytokines and bacterial products. J. Exp. Med. 182: 389–400. 23. Autenrieth, S. E., and I. B. Autenrieth. 2009. Variable antigen uptake due to different expression of the macrophage mannose receptor by dendritic cells in various inbred mouse strains. Immunology 127: 523–529. 24. Burgdorf, S., V. Lukacs-Kornek, and C. Kurts. 2006. The mannose receptor mediates uptake of soluble but not of cell-associated antigen for cross-presentation. J. Immunol. 176: 6770–6776. 25. Goodridge, H. S., D. M. Underhill, and N. Touret. 2012. Mechanisms of Fc receptor and dectin-1 activation for phagocytosis. Traffic 13: 1062–1071. 26. Mourie`s, J., G. Moron, G. Schlecht, N. Escriou, G. Dadaglio, and C. Leclerc. 2008. Plasmacytoid dendritic cells efficiently cross-prime naive T cells in vivo after TLR activation. Blood 112: 3713–3722. 27. Schlecht, G., S. Garcia, N. Escriou, A. A. Freitas, C. Leclerc, and G. Dadaglio. 2004. Murine plasmacytoid dendritic cells induce effector/memory CD8+ T-cell responses in vivo after viral stimulation. Blood 104: 1808–1815. 28. Jung, S., D. Unutmaz, P. Wong, G. Sano, K. De los Santos, T. Sparwasser, S. Wu, S. Vuthoori, K. Ko, F. Zavala, et al. 2002. In vivo depletion of CD11c+ dendritic cells abrogates priming of CD8+ T cells by exogenous cell-associated antigens. Immunity 17: 211–220. 29. Lian, Z. X., T. Okada, X. S. He, H. Kita, Y. J. Liu, A. A. Ansari, K. Kikuchi, S. Ikehara, and M. E. Gershwin. 2003. Heterogeneity of dendritic cells in the mouse liver: identification and characterization of four distinct populations. J. Immunol. 170: 2323–2330. 30. Burmeister, Y., T. Lischke, A. C. Dahler, H. W. Mages, K. P. Lam, A. J. Coyle, R. A. Kroczek, and A. Hutloff. 2008. ICOS controls the pool size of effectormemory and regulatory T cells. J. Immunol. 180: 774–782. 31. Chen, Y., S. Shen, B. K. Gorentla, J. Gao, and X. P. Zhong. 2012. Murine regulatory T cells contain hyperproliferative and death-prone subsets with differential ICOS expression. J. Immunol. 188: 1698–1707. 32. Lo¨hning, M., A. Hutloff, T. Kallinich, H. W. Mages, K. Bonhagen, A. Radbruch, E. Hamelmann, and R. A. Kroczek. 2003. Expression of ICOS in vivo defines CD4+ effector T cells with high inflammatory potential and a strong bias for secretion of interleukin 10. J. Exp. Med. 197: 181–193. 33. Fessler, J., A. Ficjan, C. Duftner, and C. Dejaco. 2013. The impact of aging on regulatory T-cells. Front. Immunol. 4: 231. 34. Nishioka, T., J. Shimizu, R. Iida, S. Yamazaki, and S. Sakaguchi. 2006. CD4+ CD25+Foxp3+ T cells and CD4+CD25-Foxp3+ T cells in aged mice. J. Immunol. 176: 6586–6593. 35. Akbari, O., G. J. Freeman, E. H. Meyer, E. A. Greenfield, T. T. Chang, A. H. Sharpe, G. Berry, R. H. DeKruyff, and D. T. Umetsu. 2002. Antigenspecific regulatory T cells develop via the ICOS-ICOS-ligand pathway and inhibit allergen-induced airway hyperreactivity. Nat. Med. 8: 1024–1032. 36. Huleatt, J. W., and L. Lefrançois. 1995. Antigen-driven induction of CD11c on intestinal intraepithelial lymphocytes and CD8+ T cells in vivo. J. Immunol. 154: 5684–5693. 37. Asselin-Labat, M. L., M. David, A. Biola-Vidamment, D. Lecoeuche, M. C. Zennaro, J. Bertoglio, and M. Pallardy. 2004. GILZ, a new target for the transcription factor FoxO3, protects T lymphocytes from interleukin-2 withdrawalinduced apoptosis. Blood 104: 215–223. 38. Cannarile, L., F. Fallarino, M. Agostini, S. Cuzzocrea, E. Mazzon, C. Vacca, T. Genovese, G. Migliorati, E. Ayroldi, and C. Riccardi. 2006. Increased GILZ expression in transgenic mice up-regulates Th-2 lymphokines. Blood 107: 1039– 1047.


This article is dedicated to the memory of Prof. Dominique Emilie, the former head of our laboratory who initiated this work. We thank H. Gary and H. Tharinger (Plaimmo, Institut Paris-Sud d’Innovation The´rapeutique/ Institut Fe´de´ratif de Recherche 141, Chatenay-Malabry, France) for excellent technical assistance. We thank M. Levant (Animex, Institut Paris-Sud d’Innovation The´rapeutique/Institut Fe´de´ratif de Recherche 141, ChatenayMalabry, France) for mice handling, and B. Choisi and M. Horckmans (INSERM UMR 996, Clamart, France) for help with genotyping. We are grateful to Drs. R. Flavell and B. Malissen for providing IL-10–GFP and Foxp3-eGFP mice, respectively. We thank J. Banchereau (INSERM U955, Eq16, Vaccine Research Institute, Cre´teil, France) for comments and critical reading of the manuscript.

5871

5872


39. Jenkins, M. K., E. Burrell, and J. D. Ashwell. 1990. Antigen presentation by resting B cells. Effectiveness at inducing T cell proliferation is determined by costimulatory signals, not T cell receptor occupancy. J. Immunol. 144: 1585– 1590. 40. Groux, H., M. Bigler, J. E. de Vries, and M. G. Roncarolo. 1996. Interleukin-10 induces a long-term antigen-specific anergic state in human CD4+ T cells. J. Exp. Med. 184: 19–29. 41. Schwartz, R. H. 1996. Models of T cell anergy: is there a common molecular mechanism? J. Exp. Med. 184: 1–8. 42. Groux, H., A. O’Garra, M. Bigler, M. Rouleau, S. Antonenko, J. E. de Vries, and M. G. Roncarolo. 1997. A CD4+ T-cell subset inhibits antigen-specific T-cell responses and prevents colitis. Nature 389: 737–742.

43. Ngo, D., E. Beaulieu, R. Gu, A. Leaney, L. Santos, H. Fan, Y. Yang, W. Kao, J. Xu, V. Escriou, et al. 2013. Divergent effects of endogenous and exogenous glucocorticoid-induced leucine zipper in animal models of inflammation and arthritis. Arthritis Rheum. 65: 1203–1212. 44. Suarez, P. E., E. G. Rodriguez, R. Soundararajan, A. M. Me´rillat, J. C. Stehle, S. Rotman, T. Roger, M. J. Voirol, J. Wang, O. Gross, et al. 2012. The glucocorticoid-induced leucine zipper (gilz/Tsc22d3-2) gene locus plays a crucial role in male fertility. Mol. Endocrinol. 26: 1000–1013. 45. Krzysiek, R. 2010. Role of glucocorticoid-induced leucine zipper (GILZ) expression by dendritic cells in tolerance induction. Transplant. Proc. 42: 3331–3332.


Enhanced glucocorticoid-induced leucine zipper in dendritic cells induces allergen-specific regulatory CD4(+) T-cells in respiratory allergies.

The Pou5f1 distal enhancer is sufficient to drive Pou5f1 promoter-EGFP expression in embryonic stem cells.

Membrane-bound metallothionein 1 of murine dendritic cells promotes the expansion of regulatory T cells in vitro.

Downregulation of the Syk Signaling Pathway in Intestinal Dendritic Cells Is Sufficient To Induce Dendritic Cells That Inhibit Colitis.

Tolerogenic Dendritic Cells for Regulatory T Cell Induction in Man.

Enhanced efferocytosis by dendritic cells underlies memory T-cell expansion and susceptibility to autoimmune disease in CD300f-deficient mice.

Mycophenolic acid-treated dendritic cells generate regulatory CD4+ T cells that suppress CD8+ T cells' allocytotoxicity.

EZH2 is crucial for both differentiation of regulatory T cells and T effector cell expansion.

Jagged-1 is required for the expansion of CD4+ CD25+ FoxP3+ regulatory T cells and tolerogenic dendritic cells by murine mesenchymal stromal cells.

Tumor-specific hepatic stellate cells (tHSCs) induces DIgR2 expression in dendritic cells to inhibit T cells.

Dual leucine zipper kinase regulates expression of axon guidance genes in mouse neuronal cells.

Activation of myeloid dendritic cells, effector cells and regulatory T cells in lichen planus.

Leptin deficiency in vivo enhances the ability of splenic dendritic cells to activate T cells.

A novel Cd8-cis-regulatory element preferentially directs expression in CD44hiCD62L+ CD8+ T cells and in CD8αα+ dendritic cells.

Memory CD4+ T cells are suppressed by CD8+ regulatory T cells in vitro and in vivo.

Low-dose intradermal infection with trypanosoma congolense leads to expansion of regulatory T cells and enhanced susceptibility to reinfection.

iPSC-Derived Regulatory Dendritic Cells Inhibit Allograft Rejection by Generating Alloantigen-Specific Regulatory T Cells.

Activated CD8+ T cells induce expansion of Vβ5+ regulatory T cells via TNFR2 signaling.

Temporal biological variability in dendritic cells and regulatory T cells in peripheral blood of healthy adults.

Plasmacytoid dendritic cells license regulatory T cells, upon iNKT-cell stimulation, to prevent autoimmune diabetes.

CD4+CXCR3+ T cells and plasmacytoid dendritic cells drive accelerated atherosclerosis associated with systemic lupus erythematosus.

developmental endothelial locus-1-dependent mechanism.

Intestine-derived Clostridium leptum induces murine tolerogenic dendritic cells and regulatory T cells in vitro.

In vitro TNF blockade enhances ex vivo expansion of regulatory T cells in patients with immune thrombocytopenia.